Reviews are the most valuable marketing asset most businesses never deliberately build. They influence whether you appear in the local pack, they are read by nearly everyone before a purchase decision, and they cost nothing to acquire except the willingness to ask.

And yet most businesses do not ask. Or they ask badly — at the wrong moment, through the wrong channel, in a way that produces nothing, and sometimes in a way that breaks Google's rules and gets the reviews they did earn removed.

This piece is about what actually works, and about the lines you should not cross even though a great many competitors are crossing them.

Why Google reviews specifically

There are many review platforms, and for most businesses only one of them really matters.

Google reviews appear directly in the search result and the map pack — the place where purchase decisions are actually made. They feed the local ranking algorithm. And they arrive attached to the account people are already signed in to, which removes the single largest source of friction: nobody has to create anything to leave one.

Which means a review on Google is worth several on a platform your customers have never heard of and would have to register for. Industry-specific sites matter in some verticals, and Trustpilot and G2 matter in others. But if you are choosing where to spend a finite amount of asking, Google is where it goes.

The moment to ask is narrower than you think

Almost every failed review campaign fails on timing.

The window is immediately after a moment of realised value. Not after the sale — after the success. The distinction matters enormously. A customer who has just bought has spent money and received nothing yet. A customer whose problem you have just solved has received the thing they wanted, and is, briefly, delighted.

That delight decays fast. Within a week it is a pleasant memory. Within a month it is baseline. Ask on day one and you are asking someone who feels good about you; ask on day thirty and you are asking someone who feels nothing in particular, and indifference produces no reviews at all.

What that means in practice, per business type:

  • Services: the day the job is finished and the customer has seen the result.
  • Retail: a few days after delivery, once they have used the thing.
  • SaaS: after a genuine success event — they hit a milestone, solved the problem they signed up for. Not on a schedule, on an event.
  • Restaurants and hospitality: that evening. Not three days later.

Ask the person who is actually happy

The counter-intuitive part: do not ask everyone.

Asking your entire customer base produces reviews from your entire customer base, including the ones who were merely satisfied, the ones who had a mediocre experience they had not thought about until you invited them to, and the ones who now remember they were annoyed about something.

Ask the customers who have shown a signal of delight — the ones who thanked you, referred someone, renewed, or came back. They are the ones who will write something warm and specific, and specificity is what makes a review persuasive.

To be explicit about the line here, because it is a real one: this is not review gating. Gating means asking everyone how they feel, and then routing only the happy ones to Google while diverting the unhappy ones to a private feedback form. That is against Google's rules, it is detectable, and it gets your reviews removed. It is also, straightforwardly, a lie about what your customers think.

The legitimate version is different: you ask the people you have good reason to believe are happy, and if one of them leaves you three stars, that review stands. You are choosing who to ask. You are not filtering what they say.

The rules you are probably breaking

Google's prohibited content policy is short and worth reading in full. The three violations businesses commit most often, usually without realising:

1. Incentivising reviews. Offering a discount, a prize draw entry, a free coffee — anything of value in exchange for a review — is prohibited. It does not matter that you did not require the review to be positive. The incentive is the violation. This is the most common breach by a distance, and "10% off your next order if you leave us a review" is a sentence that appears on thousands of receipts.

2. Review gating. As above. Screening for sentiment before deciding who gets the review link.

3. Soliciting reviews in bulk from people who are not customers. Staff, friends, family. This is easy for Google to detect — the reviewer profiles cluster, the timing clusters, the language clusters — and the penalty falls on your profile.

The consequences run from silent removal of the offending reviews to suppression of the whole profile. And the damage is not merely the lost reviews; it is that you have now built a marketing channel on a foundation that can be removed at any moment, by someone else, without warning.

What to actually say

The ask itself is usually far too elaborate. It needs three things and nothing else: a reason, a link, and an acknowledgement of the effort.

A version that works:

"Glad we got that sorted, [name]. If you have thirty seconds, a quick Google review genuinely helps people find us — most of our work comes from people who found us that way. Here's the link: [link]. And if there's anything that wasn't right, tell me instead and I'll fix it."

Why each part is there. "Glad we got that sorted" anchors it to the specific success. "Thirty seconds" sets the cost accurately and low. "Genuinely helps people find us" gives a reason that is true and is about them helping, not about you winning. The direct link removes every remaining step. And the final sentence is the honest version of gating — you are inviting criticism privately, but you are not preventing it publicly.

What to avoid: "we'd love a 5-star review" (specifying the rating is close to soliciting a false one), anything about a target or a competition, and any language that makes it feel like a transaction.

The link matters more than the wording

Every additional step between the ask and the review costs you a large fraction of your responses.

If someone has to open Google, search for your business, find the reviews section, and click "write a review", the great majority will not finish. A direct link that opens the review dialogue immediately can multiply your conversion rate several times over, and it costs nothing.

Google provides this link. Our free Google review link generator builds it from your business, and our QR code generator turns it into something you can put on a receipt, a table card, an invoice, or a van.

The QR code is disproportionately effective for physical businesses precisely because it collapses the whole journey into one action taken at the moment of delight, with the phone already in their hand.

Which channel to ask in

Ranked by how well they actually work:

  • In person, followed by a text. The highest-converting method by a wide margin. The human ask creates the intention; the text carries the link. Neither works nearly as well alone.
  • SMS. Very high open rates, very low friction, phone already in hand. Best for services and local businesses.
  • Email. Scales, works, and converts less well than SMS. Our review request email generator drafts these.
  • In-app. Good for SaaS, if genuinely triggered by a success event rather than a timer.
  • On a receipt or invoice. Passive, low conversion, effectively free. Worth doing, not worth relying on.

The pattern to notice: what works is proximity to the moment and absence of friction. Everything else is detail.

Responding to reviews, including the bad ones

Responding is not politeness. It is a signal to Google that the profile is actively managed, and it is read by every future customer who scrolls the reviews.

The bad review is the opportunity, and almost everyone wastes it. A one-star review with no response reads as an unanswered accusation. The same review with a calm, specific, non-defensive reply reads as a business that takes problems seriously — and prospective customers frequently say that this is more persuasive than the five-star reviews, because it is the only place they get to see how you behave when something goes wrong.

What works: acknowledge the specific problem, do not argue about the facts in public, say what you have done about it, and offer to continue offline. What does not: defensiveness, blaming the customer, a copy-pasted apology, or arguing. Anyone can see the whole exchange, forever.

Our review response generator drafts these, and the point is not to automate sincerity but to get past the blank page when you are angry — which is precisely when you should not be writing the reply yourself.

Volume, recency, and why the work never stops

Three things matter about your review profile, and only one is the average rating.

Volume gives credibility. Ten five-star reviews are less persuasive than two hundred at 4.6, because ten is a number a business could have manufactured and two hundred is not.

Recency matters enormously and is widely ignored. A profile whose most recent review is from 2023 suggests a business that has either stopped trading or stopped caring. Reviews decay in relevance, and a steady trickle beats a burst.

The rating matters least, within limits, and — counter-intuitively — a perfect five is slightly less persuasive than a 4.7. A flawless record reads as filtered. A handful of critical reviews, handled well, is evidence that the rest are real.

The implication is that review collection is not a campaign. It is an ongoing operational habit, wired into the moment of delivery, forever. Which is what Govarova Review Management is for: asking at the moment, in the right channel, and chasing the ones who meant to and forgot.

Turning reviews into something that works elsewhere

A review is not only a thing that sits on Google. It is customer language, written by customers, about what they actually valued — which is the most useful copy your marketing will ever have access to, and most businesses never read it.

Three things to do with it:

  • Read them for your own positioning. The words customers use to describe the benefit are better than the words your marketing team invented, because they are the ones that survived contact with a real purchase decision.
  • Put them where the decision happens. On the pricing page, the checkout, the signup form — not on a "testimonials" page nobody visits. Our star rating generator and review widget builder handle the mechanics.
  • Use them as live social proof. A recent, real review shown at the moment of hesitation is what Social Proof notifications are for.

The follow-up, and why one ask is not a system

Most people who intend to leave a review do not. Not out of malice — they meant to, and then the kettle boiled and the moment passed. The gap between intention and action is where almost all of your missing reviews live.

A single, gentle reminder recovers a large share of them. One. Not a sequence, not a campaign — one message, a few days later, in the same channel, acknowledging that they are busy and making it trivially easy.

Beyond that, stop. Chasing a review is qualitatively different from chasing a sale: you are asking someone for a favour, and pursuing someone for a favour converts goodwill into irritation faster than almost anything else you can do. The customer who was going to write something warm and now feels harassed writes nothing, or worse.

The rule that keeps this healthy: one ask, one reminder, then never again for that transaction. If they come back and buy again, that is a new moment of value and a new opportunity — but the previous silence is a closed matter, and treating it as an open one is how a review request becomes a nuisance.

What to do about a review that is unfair

It will happen. A review that describes something that did not occur, or blames you for something outside your control, or comes from someone who was never a customer.

The instinct is to fight it. Resist that for long enough to work out which of three situations you are actually in.

It violates Google's policies. Fake reviews, reviews from competitors, reviews containing personal attacks or off-topic content can be reported and are sometimes removed. The process is slow and the success rate is modest, but it exists and it costs you nothing but time.

It is unfair but genuine. The customer really did have that experience, even if you think their interpretation is wrong. This one you cannot remove and should not try to. Respond calmly, correct the factual record without arguing about the feelings, and let readers judge. A measured reply to an unreasonable review is the most persuasive thing on your entire profile, because everyone reading it has been the unreasonable customer at some point and knows it.

It is fair and you did not want to hear it. The hardest case. The correct response is to fix the thing, say publicly that you have fixed it, and move on. Businesses that do this visibly earn more from the bad review than they lost.

What never works: arguing, publishing private details about the customer to prove your case, or a lawyer's letter. All three are visible to everyone, forever, and they turn one bad review into your reputation.

Reviews and local search: the mechanical part

For a business with a physical location, reviews are not just persuasion. They are a ranking input.

Google's local ranking blends relevance, distance and prominence, and reviews feed prominence — volume, rating and recency all appear to contribute. Which means review collection is simultaneously a conversion tactic and a visibility tactic, and it is the only marketing activity that is unambiguously both.

Two practical notes that most guides miss:

  • Reviews mentioning the service by name appear to help relevance for that service. Not because you should coach people into saying magic words — that is soliciting a false review — but because a genuinely specific review naturally contains the language of the thing you actually did, which is a good reason to ask at the moment when the specifics are fresh.
  • A steady rate beats a burst. Twenty reviews in one week after two years of silence looks exactly like what it usually is, and Google has every incentive to be suspicious of it.

The dull conclusion, which is nevertheless correct: ask everyone worth asking, every time, forever, at the moment it is deserved. The compounding is the strategy.

Measuring whether any of this is working

Four numbers, and they take five minutes a month:

  • Ask rate. Of the customers who reached a genuine success moment, what fraction were actually asked? This is usually the broken number, and it is usually far lower than anyone believes — the system exists, and nobody runs it.
  • Conversion rate. Of those asked, what fraction left a review? Below about 10% and something is wrong with the channel, the link, or the timing.
  • Recency. Days since the last review. If this is creeping up, the habit has quietly stopped.
  • Response rate. What fraction of reviews got a reply? This is entirely within your control and it is visible to every prospect.

Notice that the average star rating is not on the list. It is an output, not a lever — you cannot improve it directly, only by improving the business or by cheating, and one of those is a trap. Measure the things you can actually act on. Our reputation score calculator takes the whole picture, and the NPS calculator covers the private-feedback side that never reaches Google.

Where the whole thing usually breaks

Not in the wording, and not in the channel. In the habit.

Almost every business that has a review problem has the same one: they decided to collect reviews, they asked for two weeks, they got a few, they got busy, and they stopped. Eighteen months later the most recent review is from a season nobody remembers, and the profile reads as abandoned.

The fix is not a better ask. It is to make the ask part of the process of finishing the work — the invoice, the completion email, the delivery confirmation — so that it happens whether or not anyone remembers to make it happen. Anything that depends on someone deciding to do it will stop being done, and it will stop quietly.

That is the entire argument for automating it: not that a machine writes a better request than you do, but that a machine still sends it in the week you are too busy to think about reviews at all — which is, reliably, the week you did your best work.

The part that cannot be automated

All of the above assumes something that no tool provides, and it is worth being blunt about it.

Reviews are a measurement instrument pointed at your business. If the service is mediocre, a better ask, a shorter link and a well-timed SMS will produce more reviews accurately reporting that it is mediocre. The tooling does not manufacture goodwill; it removes the friction between goodwill and its expression.

Which is a much better position than it sounds, because it means the work is legible. If you ask well, at the right moment, with no friction, and the reviews are still not good — you have learned something true and expensive about your business, and you have learned it from the people best placed to tell you.

That is worth more than the reviews.

Start this week

If you take nothing else from this: generate the direct link today, put it somewhere your team can reach in one click, and ask the next customer who thanks you. That is the entire programme, and everything else in this article is refinement on top of it.

The businesses with two hundred reviews are not doing something clever. They are doing something ordinary, every single time, for years — and the compounding does the rest. The gap between them and a business with nine reviews is almost never talent, budget, or a better tool. It is that one of them asked and the other meant to.

The reliable way to get more reviews is to ask every customer at the right moment, and to chase the ones who forget. Govarova Review Management automates the ask, Sequences handles the reminders, and you can start free.